Pastoralism: Research, Policyand Practice

Ilukor et al. Pastoralism: Research, Policy and Practice (2016) 6:14 DOI 10.1186/s13570-016-0061-1

SHORT REPORT Open Access

To eradicate or not to eradicate?Recommendations on Prosopis julifloramanagement in Afar, Ethiopia, from aninterdisciplinary perspective

John Ilukor*, Simone Rettberg, Anna Treydte and Regina Birner

Abstract

As one of the most invasive species of arid and semi-arid areas of East Africa, Prosopis juliflora has become a majorthreat to livelihoods of the Afar pastoral people and to the fragile ecosystems they live in. This paper comprisesresults from a multidisciplinary study on the spread of P. juliflora as an invasive species in the Ethiopian Afar Regionand provides suggestions on its management and control. The study investigated spread of P. juliflora, ecologicalaspects (shifts in vegetation biomass and soil properties) and socio-economic aspects (livelihood impacts, managementactivities and potential) of P. juliflora invasion. Ecological methods included conducting destructive harvesting onvegetation; soil analyses and a soil seed bank assessment as well as using allometric equations to estimate the biomassand carbon contents of P. juliflora. Socio-economic data was collected based a survey of 490 pastoral householdsincluding 213 from Amibara, 177 from Gewane and 100 from Awash Fentale. Stated and revealed preference methods(RPM) were used because impacts of P. juliflora are both tangible and intangible. The study found that wetlands (floodplains in the Awash Basin) are highly susceptible to Prosopis invasion relative to drylands. Clearing invaded land andcontinuously using it for crop farming would reduce the invasion. The study also found that the available nitrogen,phosphorus and organic carbon in the soil were high in highly invaded areas compared to less or non-invaded sites.However, the basal cover of native herbaceous vegetation and native tree diversity were found to be much reducedunder high P. juliflora-invaded areas. Results from economic analyses also reveal that the benefits of the P. juliflorainvasion in the Afar region are higher than the costs. However, some aspects such as increased risk of erosion, theimpact of P. juliflora on the water table and long-term ecological changes were not examined, thus making the totaleconomic valuation incomplete. Nonetheless, the study has captured most aspects of P. juliflora invasion in the Afarregion and concludes that sustainable management and control of P. juliflora in the Afar region may be a bettersolution than eradication.

Keywords: Prosopis juliflora, pastoralism, costs and benefits, invasive species

Background to the study and findingsThe spread of invasive plant species within the arid andsemiarid lowlands of Northeast Ethiopia is an increasingthreat for pastoral livelihoods and ecosystems. One ofthe most invasive species is Prosopis juliflora, an ever-green, fast-growing tree or shrub, native to South andCentral America (Mwangi and Swallow 2008). TheEthiopian National Biodiversity Strategy and Action

* Correspondence: john.ilukor@gmail.comMakerere University College of Agricultural and Environmental Sciences,Kampala, Uganda

© 2016 The Author(s). Open Access This articleInternational License (http://creativecommons.oreproduction in any medium, provided you givthe Creative Commons license, and indicate if

Plan (MoARD 2005) considers it to be one of the topinvasive species countrywide, next to Partheniumweed (Parthenium hysterophorus), water hyacinth(Eichhornia crassipes) and Lantana weed (Lantanacamara). P. juliflora was first introduced to the Afarregion by the Ethiopian government in the late 1970sand early 1980s to combat desertification, and by 2006,approximately 700,000 ha of land had been taken over byP. juliflora, out of which more than 70% is located in theAfar region (Admasu 2008; Ryan 2011).

is distributed under the terms of the Creative Commons Attribution 4.0rg/licenses/by/4.0/), which permits unrestricted use, distribution, ande appropriate credit to the original author(s) and the source, provide a link tochanges were made.

Ilukor et al. Pastoralism: Research, Policy and Practice (2016) 6:14 Page 2 of 8

P. juliflora has been reported to offer significant ecosys-tem and livelihood services such as microclimate regula-tion, improvement of soil fertility, reclaiming saline andalkaline soils and income and livelihood diversification(Bhojvaid and Timmer 1998; Berhanu and Tesfaye 2006).This is particularly important for the arid lands of Afar,50% of which is exposed soil, sand and rock (ANRS 2010).Along the Awash River flood plains, dense woodland andswamps have been reclaimed for irrigation (ANRS 2010).Kahi et al. (2009) found that soils under P. juliflora hadhigher organic carbon and total nitrogen than soils in theopen areas. Although soils under Acacia spp. trees hadhigher organic carbon and total nitrogen than soils underP. juliflora (Kahi et al. 2009), Acacia spp. trees are not cap-able of surviving in the Afar region. P. juliflora’s high cop-picing ability and deep roots enable them to survive indrylands where they provide a number of benefits includ-ing alternative energy sources (Ayanu et al. 2015). Thepods from P. juliflora can also be a source of nutritioushuman food (Choge et al. 2007) as well as a nutritious, lesscostly feed ingredient for livestock (Sawal et al. 2004; Steinet al. 2005; Chaturvedi and Sahoo 2013). Both crushedand uncrushed P. juliflora pods can be fed to livestock(Berhanu and Tesfaye 2006) and in pastoral areas like Afarwith acute shortage of fresh water, P. juliflora has beenseen as an alternative source of feed because it toleratesadverse arid conditions (Mahgoub et al. 2005).Recent studies by Ayanu et al. (2015) and Haregeweyn

et al. (2013) have shown that the P. juliflora invasionrate is increasing rapidly, suppressing indigenous plants,while negatively affecting human health as well as live-stock production. P. juliflora creates an environment con-ducive for mosquito breeding, limits access to wateringpoints and grazing lands and offers cover for wild preda-tory animals such as lions (Panthera leo) (Pasiecznik et al.2001; Mehari 2015). Attempts to access these areas haveresulted in injuries to both animals and humans (Mwangiand Swallow 2008). Afar pastoralists who call it the “DevilTree” and the “AIDS” to the animals want it to be eradi-cated. Policy-makers and development partners are facedwith the dilemma of whether to completely eradicate P.juliflora or manage this species intensely. However, themeasures that have been adopted in the Afar region havethus far not been able to manage and control this species’rapid spread and it appears that the negative impactsexceed the positive impacts, especially for pastoralists, themain inhabitants of Afar Region in Ethiopia.Against this background, the German Ministry for Eco-

nomic Cooperation and GIZ (Gesellschaft für Internation-ale Zusammenarbeit) decided to fund an interdisciplinaryresearch study on “Woody Encroachment in the AfarRegion, Ethiopia: Impact Assessment of Prosopis Invasionand Participative Management Approaches”. In 2013 to14, the study was implemented by Ethiopian-German

research teams led by the University of Hohenheim (ecol-ogy, economy) and the University of Bonn (sociology,remote sensing) in association with the Mekelle Univer-sity, Samara University and Ethiopian Environment andForestry Research Institute. The project’s objective was toassess the extent and impact of land degradation throughP. juliflora in the Afar Region, and to estimate the costsand benefits of P. juliflora invasion. It was decided tofocus on an empirical case study within one of the mostheavily invaded areas in Afar, the Middle Awash Basin, theseasonally inundated floodplains of Gewane and of Ami-bara Woreda, where P. juliflora has spread rapidly overthe last decades (Müller-Mahn and Rettberg 2012).

ResultsThe invasion and spread of P. julifloraThe quantitative assessment of P. juliflora spread withinthe Baadu area (Gewane Woreda) was based on a compari-son of Cloud-free Landsat ETM (30-m resolution) andASTER (15-m resolution) satellite images taken during thedry seasons (October–March) of 2000, 2005, 2010, and2013. The results revealed that the rate of P. juliflora inva-sion over the past two decades is tremendously high, andwithin five years (2000 to 2005), the total area of wetlandsinvaded by P. juliflora increased from 3,600 to 8, 000 ha(i.e. from 8% to 18%). About 30% of the wetlands were in-vaded in 2010 - which is equivalent to a total area over13,000 ha - and by the end of 2013, almost 40% of the wet-lands in Baadu were invaded. This implies a 10% incrementover a period of three years when compared with 2010(Ayanu et al. 2015). The area of agricultural land invadedby P. juliflora was only 2 ha (<1%) even though it increasedto 76 ha (4%) in the year 2005. By 2010, the total agricul-tural land invaded by P. juliflora had increased to 166 ha,but this amounted only to 2% of the total agricultural land.The reduction in the percentage of the invaded agriculturalland is attributed to increased cultivated area comparedwith previous years, and overall increase in cultivated landcould also be associated with P. juliflora, which is known toreclaim saline and alkaline soils, which might have madethe soils available for cultivation. By the end of 2013, 4%(327 ha) of total area of agricultural land was invaded, andthis is associated with the fact that investors abandonedtheir farmland, allowing P. juliflora to continue invading.Only 1% of the total drylands were invaded, and this isprobably because animals do not graze in these areas andthe soils have less moisture available to facilitate P. juliflorainvasion (Treydte et al. 2014; Ayanu et al. 2015).

Ecological consequences of P. juliflora invasionThe acacia trees are not capable of surviving in Afar re-gion and continue to provide fuel wood, charcoal andregulate the microclimate compared to prosopis. This

Ilukor et al. Pastoralism: Research, Policy and Practice (2016) 6:14 Page 3 of 8

trend was similar for above and below ground biomass(see Figure 1).The study also found that the available phosphorus

and organic carbon in the soil were more than twice ashigh at highly and intermediate infested sites comparedto low and non-infested sites. The nitrogen level and themoisture content in soils of highly infested areas were23% and 11% to 13%, respectively. These levels are morethan twice as high compared to those of low or non-infested areas. However, species diversity and richness ofthe woody vegetation in Gewane and Amibara werefound to be very low (four to five species) even in areaswithout P. juliflora. In areas of high and medium P. juli-flora invasion, the individuals or plants of other specieswere not available or recorded (see Table 1). Therefore,P. juliflora might reduce this diversity further. The basalcover of native herbaceous vegetation was found to bemuch reduced under high P. juliflora infestation.However, the seed bank of soils under high P. juliflora

infestation showed up to 14 different species availablewhen grown in the greenhouse. Hence, there is potentialfor P. juliflora-infested soils that, once cleared, nativevegetation might recolonize certain areas when resprout-ing of P. juliflora is suppressed simultaneously. Further,the number of P. juliflora seedlings was proportionallylower the higher the invasion rate, indicating that P.juliflora might suppress seedling growth through allelop-athy and, hence, simulate self-thinning in areas of highpresence.With the loss of indigenous grasses and other plant spe-

cies, the main fodder resource for grazers such as cattle,the number and productivity of animals has reduced sub-stantially (Hamedu 2014). Yosef et al. (2013) using datafrom the Central Statistical Agency (CSA) of Ethiopiademonstrated that cattle and camel populations in theAmibara zone, which was the study area, declined at a rate

Figure 1 Above and below ground and total biomass of Prosopis juliflorathrough destructive harvesting

of 36% and 20%, respectively, between 1997 and 2011(see Table 2). Although the population of cattle inalmost all other zones is declining, the percentage de-cline in Amibara is the highest compared to other zones(Yosef et al. 2013). In addition, Amibara is the only zonefor which both camel and cattle number are declining andis also the zone which is highly invaded with P. juliflora.Further, socio-economic impacts triggered by the inva-

sion of P. juliflora comprised increased health risks dueto higher exposure to predators and malaria, constrainedaccess to water points and the emergence of new fatalanimal diseases like “Harmaku”, resulting from cyto-toxins damaging the neurons of the intoxicated animals(Silva et al. 2007).Nevertheless, P. juliflora is only one among several

ecological problems. A ranking exercise within kebeles(A Kebele is the smallest administrative unit of Ethiopiasimilar to a ward) affected by P. juliflora revealed thatAfar pastoralists perceived the diminishing water volumeof the Awash River to be as problematic as the spread ofP. juliflora or even worse (Hamedu 2014). Pasieczniket al. (2001) contend that diminishing water volumes inrivers and water courses can be attributed to P. juliflorain areas invaded by Prosopis.Most pastoralists within Baadu, previously known for

their wealth due to large cattle herds, now live underconditions of chronic food insecurity (Rettberg 2010).The number of livestock is declining as households areselling off animals to buy food, and about 64% of therural Afar households are consuming three or fewerfood groups out of seven. In Ethiopia, Afar State has thehouseholds with the highest food expenditure and mal-nutrition as well as the lowest household cereal stockper capita (1 kg/person) at any given time (WFP 2009;CSA 2014). To cope with this food insecurity challenge,there has been a significant diversification of pastoralist

-invaded sites across the different infestation categories as quantified

Table 1 Number of individuals per species for all woody plantsrecorded in the sample quadrants across the different Prosopisjuliflora encroachment sites

Site Species High Medium Low None

Amibara Acacia mellifera 0 9 25 0

Acacia senegal 0 0 0 928

Dobera glabra 50 22 3 0

Prosopis juliflora 4200 503 325 44

Total 4250 534 353 972

Gewane Acacia mellifera 0 0 0 13

Acacia senegal 0 0 0 863

Acacia seyal 0 0 25 0

Acacia spp. 0 0 50 0

Acacia tortilis 0 0 0 13

Balanites aegyptiaca 0 0 13 0

Dobera glabra 0 0 0 12

Prosopis juliflora 3850 1775 1513 13

Total 3850 1775 1600 913

Ilukor et al. Pastoralism: Research, Policy and Practice (2016) 6:14 Page 4 of 8

livelihood strategies and engagement into non-pastoraleconomic activities. More and more pastoralists becamesedentary within the last 20 years, performing differentkinds of income-generating activities (wage labour, pettytrading and sale of charcoal, firewood, grass mats, etc.)and small-scale irrigation agriculture (Rettberg 2010;Müller-Mahn and Rettberg 2012). Still, in most seden-tary households, there are some boys and men who keepon moving with the remaining livestock and livestock isstill the main food security asset.

Social and economic impacts of P. julifloraSocio-economic data was collected from a survey of490 pastoral households including 213 from Amibara,177 from Gewane and 100 from Awash Fentale. Thetotal economic value approach was used, and the valu-ation techniques applied are a combination of statedand revealed preference methods. Since the impacts ofP. juliflora are both tangible and intangible, combiningrevealed and stated preference methods ensures thatmeasurements of prices are drawn from the actual finan-cial setting and welfare measurements are less prone tobias (Adamowicz et al. 1994, 1997; Earnhart 2001). Inusing stated preference techniques, households were

Table 2 The percentage change of cattle and camel in Afar State (ABorana zone

Zones

Shinile Jijiga

Cattle Camel Cattle Camel

Percentage change −14.6 16.5 −8.21 241

Source: Yosef et al. (2013)

asked to subjectively state the quantities and the value ofenvironmental goods generated from Prosopis as shownbelow. In addition, pastoralists were asked to state thenumber of animals affected in the previous year, numberof animals that died or were slaughtered, and to estimatethe loss in weight as well as milk production that theyattributed to the invasion of P. juliflora. This informationwas collected for each of six categories of livestock. Forproducts that are traded in the market like milk and meatweights, the revealed preference method was used.Only few Afar people value P. juliflora for its eco-

nomic potential, considering it to be a “black gold” dueto its monetary benefits through charcoal productionand trade (Datona 2014). The results in Table 3 showthat the main local use of Prosopis is for fuel wood,house construction and fencing houses and kraals. Veryfew households are engaged in selling fuel wood or char-coal burning. Also, very few households feed animalswith pods and Prosopis leaves. This could explain whymost households preferred complete eradication of theProsopis. On average, 46% of the environmental incomeis associated with Prosopis. Only 18%, mainly young andeducated Afar households, have been able to generatehigh profits through charcoal trade but with severe eco-logical consequences as indigenous trees are similarly(but illegally) cut, which is against the traditional law ofthe Afar (maada). This law is not respected anymore andthe role of customary institutions in natural resourcemanagement is increasingly undermined, driven by thedesire to diversify livelihoods and income sources inselling wood and charcoal (Datona 2014). There is alsoserious concern by some interviewed pastoralists that thefew powerful people engaged in charcoal making are infavour of the spread of P. juliflora and, hence, are actingagainst encroachment control efforts. The risk ofresource-based land conflicts within and among Afar clansand neighbouring ethnic groups in the P. juliflora-affectedareas is increasing with the commodification of landthrough charcoal trade and agriculture (Hamedu 2014).Results from economic analyses also reveal that the

benefits of the P. juliflora invasion in the Afar region arehigher than the costs. The total benefits in EthiopianBirr amounted to about 4.4 billion (231 million USD),and the costs are estimated to be about 2.2 billion ETB(116 million USD). Results also show that if the currentinvasion level is not controlled, the net present value in

mibara and Mille), Somali region states (Shinile and Jijiga) and

Amibara Mille Borena

Cattle Camel Cattle Camel Cattle Camel

−36.2 −20 332 1203 −30.7 65.9

Table 3 Main uses of Prosopis juliflora among Afar pastoral households

Environmental good Harvests/year Value/year (ETB) Proportion due to Prosopis % of HHs benefitting Prosopis income/hh

Fuelwood, home consumption 119 2983 65 96 1939

Fuel wood for sale 153 5648 76 3 4292

Charcoal, home consumption (bags) 35 1712 86 18 1472

Charcoal for sale (bags) 1043 57,631 90 18 51,868

Poles for house construction 69 1664 54 52 899

Poles for house repair 26 210 68 47 143

Fencing (homes farm, kraal) 34 699 71 74 496

Farm implements 14 261 43 21 112

Household furniture 77 1636 17 4 278

Honey (litres) 48 5760 5 0.2 288

Wild fruits 65 530 4 3 21

Bush meat 12 2550 23 1 587

Medicinal use 49 1294 6 10 78

Leaves for livestock feed 175 3995 30 12 1199

Pods for livestock feed 136 3374 87 39 2935

Total 95,890 46 69,248

Source: Field Survey Data (January 2014 and December 2013): The harvest per year, value per year, proportion and the proportion of households earning a givenincome are generated from household survey data. Income due to Prosopis is computed from value per year based on proportion income associated to Prosopis

Ilukor et al. Pastoralism: Research, Policy and Practice (2016) 6:14 Page 5 of 8

the next 30 years would be negative (37 trillion Birr)(Ilukor et al. 2014). However, the benefits and costs tendto vary between user groups such as mobile pastoralists,sedentary small-scale agro-pastoralists and large-scalefarmers (Ayanu et al. 2015). In this study, we found thatthe benefits to sedentary small-scale agro-pastoralists whoparticipate in wood and charcoal trade as well as to large-scale crop farmers are higher than the costs because landis cultivated and the wood is sold. The net present valuefor households who cut prosopis, sell its wood or charcoalis 516 ETB (27 USD) and for those who cultivate land andalso sell wood or charcoal is 7985 ETB (420 USD) peryear. Nevertheless, if the invasion is not controlled, thebenefits to these user groups would be less than the costsin the near future. For this reason, pastoralists prefercomplete eradication and are willing to contribute cash aswell as labour to reach this goal relative to non-pastoralhouseholds (Tilahun et al. 2014). Our results also revealedthat pastoralists are willing to pay 246 ETB (13 USD)and 38 labour days per household in a year to eradi-cate P. juliflora.However, complete eradication might have significant

impacts on the environment in the form of increasedwind erosion, evapotranspiration, and sun radiation ofopened areas (Haregeweyn et al. 2013). It would furtherreduce the ability of the Afar environment to sequestercarbon as the estimated total C stock density derivedfrom the P. juliflora invasion was about 40% higher(86 Mg C ha−1) at high than at low (50 Mg C ha−1)invasion categories (p < 0.05). Analyses of existing data

about P. juliflora invasion generated from a study byHaregeweyn et al.(2013) and climate data acquired fromthe Amibara weather station reveal that the P. juliflorainvasion has been useful in moderating climate variablesand reducing expansion of desertification in the Afarregion (Ilukor et al. 2014). Increased plant biomassabove and below ground as well as higher organic mattercontent in soils through P. juliflora invasion can contrib-ute to capturing CO2 from the atmosphere, which isimportant for climate change mitigation (Hasen-Yusufet al. 2013). It further can promote future Carbon tradeefforts, which might provide an alternative income gen-eration through payment for environmental services(PES). Moreover, as results from the ecological analysesreveal, the available phosphorus, nitrogen and carbon inthe soil were found to be high in highly and intermediateP. juliflora-infested sites. Therefore, the soil propertiesin highly infested sites are suitable for plant growth orcrop production once P. juliflora has been removed.Based on the soil seed bank study, the ecosystem mightrecover from P. juliflora infestation and native speciesmight return. However, care has to be taken to preventreinvasion of cleared sites as well as overgrazing as thesefactors threaten the competitive ability of the nativeplants in this semi-arid region.

Conclusion and recommendationsThe results of the joint impact assessment presentedabove suggested that there is need for a participative,institutionalized, locally owned and government-led

Ilukor et al. Pastoralism: Research, Policy and Practice (2016) 6:14 Page 6 of 8

management approach. Although this study tried to cap-ture most aspects of the P. juliflora invasion in the Afar re-gion, some aspects such as increased risk of erosion, theimpact of P. juliflora on the water table and long-termecological changes were not examined, making the totaleconomic evaluation incomplete. However, based on ourresults, we provide the following recommendations forthe sustainable management and control of P. juliflora inthe Afar region.

Securing sustainable land reclamationResults from the joint study suggest that total eradica-tion of P. juliflora is impossible and not sustainable. Theefforts and resources should be directed to clearing ofP. juliflora in some selected areas with use of the woodfor either charcoal or fuel wood and (b) immediateutilization of the cleared land for either pasture or cropfarming with the objective of diversifying livelihoodstrategies. Implementation of this recommendation willrequire creating innovative ways of involving pastoral-ists in the mechanical clearing of P. juliflora by tappinginto local knowledge, local institutions and providingland use rights of the cleared land as well as suitabletools for farming, especially generators for irrigation.Pastoral clans, institutions and locations should bemapped and documented. Each clan should be empow-ered to oversee the clearing of P. juliflora. The clansshould be supported (financially and technically) andcommunity rules should be revised to identify weak-nesses and strengths in the management of P. juliflora.This support should essentially come from the govern-ment and even development partners. In addition, thesame clans should also be offered secure rights to usethe land for crop or forage production. Secure rightsare important because currently the land belongs to thegovernment, and households who were members of thecooperative noted that the land they cleared off P.juliflora was given away to the investor. This loss in’ownership’ certainly is one of disincentives to the man-agement of P. juliflora in Afar.

Building multi-level institutionsFor any management approach to become effective onthe ground, there has to be a strong institutional struc-ture with clear mandates and responsibilities. Thereshould be a clear leadership in the control of P. juliflora.One organization at the federal level (preferably theMinistry of Agriculture) should be given the mandate asa lead organization to control the spread of P. juliflora(Chekol and Neumann 2014). Under the lead organization,further institutional structures or arrangements shouldbe developed at all administrative levels at the regionallevel. All these governmental institutional structureswill need transparent mandates and sufficient budgets

for P. juliflora management. Local institutions at theregional and Woreda levels should be linked to cus-tomary institutional structures of natural resourcemanagement, e.g. through the establishment of jointboards or management committees.

Strengthening local institutions through participation andownershipBoth local- and national-level institutions and arrange-ments are needed to ensure sustainable management of P.juliflora. The local P. juliflora management institutionsshould be developed in line with the existing local institu-tions on pasture management and should be recognizedby federal and regional governments. In Afar, there arecustomary leaders called makabantu that are involved indecision-making in matters related to clan grazing lands,relationships with other clans, neighbours and the stateand the fiimat abba who enforce decisions made by themakabantu of individual clans and the clan leaders fromthe associated grazing lands ulooto) (see Oba 2012).These existing local institutional structures should be

activated and involved in the management of Prosopis,and this would require documenting existing ‘clans’ andan examination of their community rules before adapt-ing them to the management of P. juliflora. In addition,this process would ensure that P. juliflora managementinstitutions are developed participatively by involvingcommunities and their leaders. Participatively developedinstitutions that are locally owned are the key to sustainablemanagement and control of P. juliflora in the Afar region.They will ensure that pastoralist knowledge and views formthe basis for control and management of P. juliflora, thusminimizing the tendency of forcing interventions on them.It also minimizes the “free rider problem”, the tendency ofsome pastoralists or households not to participate in P. juli-flora clearing or management but to benefit from the clear-ing by others. By providing social sanctions and incentives,pastoralists themselves who are strongly affected by P. juli-flora invasion would be more actively involved in its man-agement and control other than armed forces that have noincentives. Group-based sanctions or incentives such asclan sanctions or multi-clan sanctions would even bemore effective (Reuben 2003).

Focusing on specific P. juliflora-invaded areasEcologically, some areas are better suited for P. julifloraeradication than others. Areas with high and dense P.juliflora cover might be too difficult and labour/financeconsuming. These areas are likely to decline in P. juli-flora in the future, anyway, due to self-thinning effects.In contrast, areas that have been invaded rather recentlyshould be cleared selectively and intensely (includingroot removal). The latter areas are less densely popu-lated with P. juliflora and might show the positive effects

Ilukor et al. Pastoralism: Research, Policy and Practice (2016) 6:14 Page 7 of 8

of this species onto the soil nutrient and moisture prop-erties as found in our study. These positive soil charac-teristics can be used for consecutive planting of nativespecies or cropping fields.The cleared areas should immediately be restocked with

native woody and grassy vegetation to prevent new inva-sions as abandoned and unused areas are preferredcolonization grounds for P. juliflora. Hence, restocked na-tive woody and grassy vegetation or a cropping field, e.g. adense maize field, will prevent P. juliflora from spreading.Further, the remaining grazing areas should be managed

such that no overgrazing effects arise as the latter willencourage P. juliflora invasions. The grass layer must bekept intact and should be interspersed with few native treesbut P. juliflora growing into tree shape can also be allowed.The spread of P. juliflora should be undermined by

collecting pods regularly and crushing them before live-stock can forage on these plant parts as livestock havebeen shown to be an important dispersal vector. Alter-natively, livestock faeces could be collected in the courseof a day and used as fuel for fires, once dried, or asmanure for crop fields in the near surroundings, after P.juliflora pods have been removed.

Flexibility of management interventionsP. juliflora control or management interventions shouldbe context-specific, culturally sensitive and should ac-count for the interest of both mobile pastoralism andagro-pastoralism. Clans currently affected by the P. juli-flora invasion in proximity to the Awash River combinedifferent livelihood strategies, including mobile pastoral-ism, settled small-scale agriculture and charcoal produc-tion. The reclamation and future use of land has to takenote of this diversity, which should be supported. Inaddition, there is need to improve the provision of ani-mal and human health care services in the Afar regionto be able to handle new animal diseases like Harmakuand increase d incidence in malaria.

Strengthening regional cooperation and learningIn addition to local and national institutions, there is aneed for regional policies and institutions to facilitateefforts to control and manage the negative impacts of P.juliflora and optimize the positive impacts of this speciesthrough sharing knowledge, advocacy and research.IGAD is an Eastern Africa organization created in 1996to replace the Intergovernmental Authority on Droughtand Development (IGADD) which was founded in 1986.The organization was to help member states Djibouti,Ethiopia, Kenya, Somalia, Sudan and Uganda to cope upwith the recurring and severe droughts and other naturaldisasters between 1974 and 1984 which caused wide-spread famine, ecological degradation and economichardship in the Eastern Africa region (AU 1996). The

establishment of the regional institutions and policiesshould be spearheaded by the Intergovernmental Authorityon Development (IGAD).

Encouraging further researchThere is a need to invest in P. juliflora research in Afarin order to develop innovative ways of controlling thisspecies but also assessing the impact of P. juliflorainvasion in the Afar region of Ethiopia. Particular areasof research that need to be addressed are the following:

(a)The feasibility of biological control agents usingexperiences from other countries.

(b)The relationship between spreading patterns andhydrological conditions.

(c)The role of P. juliflora in desalinization and cropproduction.

(d)An in-depth institutional analysis to highlight whythe existing P. juliflora management strategy forEthiopia has not been implemented so far and whatinstitutional arrangements will work for Prosopismanagement in Afar.

(e)Research on innovative uses of P. juliflora thatare socially acceptable and suited to the Afarregion and people, for example, on the possibleuse of P. juliflora pods as a source of humannutritious foods, medicinal uses and as the basisof livelihood diversification, based on theexperience from other countries.

Authors’ contributionsJI was leading the development of the synthesis and the economic analysis,SR was in charge of synthesizing the social analysis and ACT was in chargeof synthesizing the ecological analysis. RB was the project head, and criticallyreviewed the results and recommendations. All authors read and approvedthe final manuscript.

Authors’ informationJI is a lecturer at Makerere University and Research Associate consultant forFAO based at International Institute of Tropical Agriculture, Malawi. Hisresearch interest is on application of New Institutional Economics to analysefood security, service delivery and environmental challenges in developingcountries. SR is a social scientist whose research interest is in GeographicalDevelopment Studies in Africa.AT is an agroecologist and scientific assistant at Agro-ecology in the Tropicsand Subtropics, University of Hohenheim, Germany. RB is the Chair of ‘Socialand Institutional Change in Agricultural Development’ at the University ofHohenheim, Germany, and one of her research interests is on how ruralservices, including livestock services, can be delivered more effectively torural poor/marginalized people.

Competing interestsThe authors declare that they have no competing interests.

Received: 10 October 2015 Accepted: 4 July 2016

ReferencesAdamowicz, W., J. Louviere, and M. Williams. 1994. Combining revealed

and stated preference methods for valuing environmental amenities.Journal of Environmental Economics and Management 26: 271–292.doi:10.1006/jeem.1994.1017.

Ilukor et al. Pastoralism: Research, Policy and Practice (2016) 6:14 Page 8 of 8

Adamowicz, W., Joffre Swait, Peter Boxall, Jordan Louviere, and Michael Williams.1997. Perceptions vs objective measures of environmental quality incombined revealed-stated preference models of environmental valuation.Journal of Environmental and Management 32: 65–84.

Admasu, D. 2008. Invasive plants and food security: The case of Prosopis juliflora inthe Afar region of Ethiopia, 1–13. FARM-Africa, IUCN.

ANRS. 2010. Afar National Regional State Programme of Plan on Adaptation toClimate Change. Environmental Protection Authority (EPA) of the FederalDemocratic Republic of Ethiopia.

AU. 1996. Intergovernmental Authority for Development (IGAD). http://au.int/en/recs/igad.

Ayanu, Yohannes, Anke Jentsch, and Detlef Müller-Mahn. 2015. Ecosystemengineer unleashed: Prosopis juliflora threatening ecosystem services?Regional Environmental Change 15: 155–167. doi:10.1007/s10113-014-0616-x.

Berhanu, Abiyot, and Getachew Tesfaye. 2006. The Prosopis dilemma, impacts ondryland biodiversity and some controlling methods. Journal of the Drylands1: 158–164.

Bhojvaid, P.P., and V.R. Timmer. 1998. Soil dynamics in an age sequence ofProsopis juliflora planted for sodic soil restoration in India. Forest Ecology andManagement 106: 181–193. doi:10.1016/S0378-1127(97)00310-1.

Chaturvedi, Om Hari, and Artabandhu Sahoo. 2013. Nutrient utilization andrumen metabolism in sheep fed Prosopis juliflora pods and Cenchrus grass.SpringerPlus 2: 598. doi:10.1186/2193-1801-2-598.

Chekol, Wondimagegne, and Irmfried Neumann. 2014. Prosopis juliflora,Parthenium and beyond, challenges for an integrated strategy of IAS controlin the Afar Region. In Managing Prosopis Juliflora for better (agro-) pastoralLivelihoods in the Horn of Africa.

Choge, S.K., N.M. Pasiecznik, M. Harvey, J. Wright, S.Z. Awan, and P.J.C. Harris.2007. Prosopis pods as human food, with special reference to Kenya. InWater SA 33: 419–424. doi:10.4314/wsa.v33i3.49162.

CSA. 2014. Ethiopia comprehensive food security and vulnerability analysis. AddisAbaba.

Datona, Mohammed. 2014. Socio-economic impacts of Prosopis juliflora-relatedcharcoal trade in Gewane Woreda, Afar Region. In Managing Prosopis Juliflorafor better (agro -) pastoral Livelihoods in the Horn of Africa.

Earnhart, Dietrich. 2001. Combining revealed and stated preference methodsto value environmental amenities at residential locations. Land Economics77: 12–29. doi:http://le.uwpress.org/content/77/1/12.full.pdf+html.

Hamedu, Herrie. 2014. Socioeconomic and ecological impacts of Prosopis juliflorainvasion in Gewane and Buremudaytu Woredas of the Afar region. InRegional conference on managing Prosopis juliflora for better (agro-)pastorallivelihoods in the Horn of Africa, 118–124. Addis Ababa, Ethiopia: DesalegnHotel.

Haregeweyn, N., A. Tsunekawa, M. Tsubo, D. Meshesha, and A. Melkie. 2013.Analysis of the invasion rate, impacts and control measures of Prosopisjuliflora: A case study of Amibara District, Eastern Ethiopia. EnvironmentalMonitoring and Assessment 185: 7527–42. doi:10.1007/s10661-013-3117-3.

Hasen-Yusuf, M., A.C. Treydte, E. Abule, and J. Sauerborn. 2013. Predictingaboveground biomass of woody encroacher species in semi-arid rangelands,Ethiopia. Journal of Arid Environments 96: 64–72.

Ilukor, John, Regina Birner, Mesfin Tilahun, and Shimelis Getu. 2014. A social-economic assessment of the impact of Prosopis juliflora invasion andparticipative management approaches in the Afar Region, Ethiopia. InManaging Prosopis Juliflora for better (agro-) pastoral Livelihoods in the Horn ofAfrica Proceedings of the Regional Conference May 1 - May 2, 2014, AddisAbaba, Ethiopia.

Kahi, C.H., R.K. Ngugi, S.M. Mureithi, and J.C. Ng’ethe. 2009. The canopy effects ofProsopis juliflora (dc.) and Acacia tortilis (hayne) trees on herbaceous plantsspecies and soil physico-chemical properties in Njemps flats, Kenya. Tropicaland Subtropical Agroecosystems 1: 117–122.

Mahgoub, Osman, Isam T. Kadim, Neil E. Forsberg, Dawood S. Al-Ajmi, Naseeb M.Al-Saqry, Abdullah S. Al-Abri, and Kanthi Annamalai. 2005. Evaluation ofMeskit (Prosopis juliflora) pods as a feed for goats. Animal Feed Science andTechnology 121: 319–327. doi:10.1016/j.anifeedsci.2005.01.016.

Mehari, Zeraye H. 2015. The invasion of Prosopis juliflora and Afar pastorallivelihoods in the Middle Awash area of Ethiopia. Ecological Processes 4.Ecological Processes: 1–13. doi:10.1186/s13717-015-0039-8.

MoARD. 2005. National biodiversity strategy and action plan. Addis Ababa:Government of the federal democratic republic of Ethiopia.

Müller-Mahn, D., and S. Rettberg. 2012. Human-environment interactions: theinvasion of Prosopis juliflora in the drylands of northeast Ethiopia. Cambridge:Changing deserts Whitehorse Press.

Mwangi, Esther, and Brent Swallow. 2008. Prosopis juliflora invasion and rurallivelihoods in the Lake Baringo area of Kenya. Conservation and Society6: 130–140.

Oba, Gufu. 2012. Harnessing pastoralists’ indigenous knowledge for rangelandmanagement: Three African case studies. Pastoralism: Research, Policy andPractice 2: 1. doi:10.1186/2041-7136-2-1.

Pasiecznik, P Felker, PJC Harris, LN Harsh, JC Tewari, G Cruz, K Cadoret, and LJMaldonado. 2001. The Prosopis juliflora - Prosopis pallida complex: Amonograph.

Rettberg, Simone. 2010. Contested narratives of pastoral vulnerability and risk inEthiopia’ s Afar region 1. doi:10.3362/2041-7136.2010.014.

Reuben, Ernesto. 2003. The evolution of theories of collective action. Amsterdan:Universidad de Amsterdam.

Ryan, Faith. 2011. US Forest Service Technical Assistance Trip to Ethiopia: Invasivespecies management. Report Submitted to USAID.

Sawal, R. K., Ram Ratan, and S. B S Yadav. 2004. Mesquite (Prosopis juliflora) podsas a feed resource for livestock - a review. Asian-Australasian Journal ofAnimal Sciences.

Silva, A.M., A.R. Silva, A.M. Pinheiro, S.R. Freitas, V.D. Silva, C.S. Souza, J.B. Hughes,et al. 2007. Alkaloids from Prosopis juliflora leaves induce glialactivation, cytotoxicity and stimulate NO production. Toxicon 49: 601–14.doi:10.1016/j.toxicon.2006.07.037.

Stein, R B D, L R a de Toledo, F Q de Almeida, a C Arnaut, L T Patitucci, J S Neto,and V T M da Costa. 2005. Effects of feeding mesquite pod meal (Prosopisjuliflora (Swartz) DC) for horses. Revista Brasileira De Zootecnia-BrazilianJournal of Animal Science 34:1240–1247.

Tilahun, Mesfin, Regina Birner, and John Ilukor. 2014. Households’ demand formitigation of Prosopis juliflora invasion in the Afar Region of Ethiopia: Acontingent valuation. In Managing Prosopis Juliflora for better (agro-) pastoralLivelihoods in the Horn of Africa Proceedings of the Regional Conference May 1- May 2, 2014, Addis Ababa, Ethiopia.

Treydte, Anna C., Emiru Birhane, and Abeje Eshete. 2014. Ecological challengesand potential carbon storage benefits of Prosopis juliflora in Afar. InManaging Prosopis Juliflora for better (agro-) pastoral Livelihoods in the Horn ofAfrica Proceedings of the Regional Conference May 1 - May 2, 2014, AddisAbaba, Ethiopia.

WFP. 2009. Food security and vulnerability in selected towns of Afar and AmharaRegions, Ethiopia: WFP-Ethiopia Vulnerability Assessment and Mapping (VAM).Addis Ababa, Ethiopia.

Yosef, T., U. Mengistu, A. Solomon, Y.K. Mohammed, and K. Kefelegn. 2013. Cameland cattle population dynamics and livelihood diversification as a responseto climate change in pastoral areas of Ethiopia. Livestock Research for RuralDevelopment 25: 1–10.

Submit your manuscript to a journal and benefi t from:

7 Convenient online submission

7 Rigorous peer review

7 Immediate publication on acceptance

7 Open access: articles freely available online

7 High visibility within the fi eld

7 Retaining the copyright to your article

Submit your next manuscript at 7 springeropen.com